Article

CMRP Formula Sheet: Key Metrics by Pillar, With Examples

· Monarch Reliability

You can't bring a formula sheet into the CMRP exam. SMRPCO's Candidate Handbook describes it as a closed-book exam: no notes or reference materials (unless SMRPCO has authorized an accommodation), with an on-screen calculator and no outside calculators. So whatever formulas you need, you have to carry in your head.

The good news: most of the math is a small set of ratios, mostly SMRP's numbered maintenance and reliability metrics, and nearly all of them are "this bucket of hours (or dollars) divided by that bucket." People rarely miss these because the arithmetic is hard. They miss them because they pick the wrong denominator.

Below is our formula sheet, grouped by the five CMRP pillars, plus a worked example that runs one plant's month through the key formulas. For the full set of numbered SMRP metrics, each with its equation, variable key, and edition notes, see our SMRP formulas page.

How to use this sheet

  • Read the denominator first. Uptime and availability have the same numerator and different denominators. Rework is divided by all maintenance hours, not just corrective hours.
  • Follow the question. Where SMRP's Best Practices (7th edition) is inconsistent, we say so below. On the exam, use the definition the question gives you.

The time model (Pillars 2 and 3)

These buckets feed OEE, TEEP, availability, and uptime.

Symbol Bucket Formula or meaning
T_a Total available time Every hour in the period (720 h in a 30-day month)
T_i Idle time Hours the asset wasn't needed: no demand, no feedstock, no raw materials, administrative shutdown. Not downtime.
T_d Total downtime Scheduled downtime + unscheduled downtime
T_o Uptime T_a − (T_i + T_d)
T_u Utilization time T_a − T_i (the hours the asset was scheduled)

Pillar 1: Business and Management

Metric Formula
Maintenance cost as % of RAV (Total maintenance cost ÷ Replacement asset value) × 100
Stocked MRO inventory as % of RAV (Stocked MRO inventory value ÷ RAV) × 100
RAV per craft-wage employee RAV ÷ Craft-wage headcount
Maintenance unit cost Total maintenance cost ÷ Units produced

RAV means replacement asset value: what it would cost to replace the asset today. It isn't book value or insured value.

Pillar 2: Manufacturing Process Reliability

Metric Formula
Availability A = T_o ÷ (T_a − T_i)
Uptime % (T_o ÷ T_a) × 100
Utilization % (T_u ÷ T_a) × 100
Performance P = Actual rate ÷ Best rate, or Best cycle time ÷ Actual cycle time
Quality rate Q = Good units ÷ Total units
OEE A × P × Q
TEEP U × A × P × Q

Two notes. The 7th edition calculates availability net of idle time, so a line with no weekend orders isn't "unavailable." And its OEE metric box appears to print the cycle-time ratio upside down (actual ÷ best). The OEE guideline in the same book uses best ÷ actual, which keeps performance at or below 100%, and so do we. Our OEE formula page covers this in detail.

Pillar 3: Equipment Reliability

Metric Formula Applies to
MTBF Operating time ÷ Number of failures Repairable items
MTTF Operating time to failure ÷ Number of items run to failure Non-repairable items
MTTR Total repair time ÷ Number of repairs Repair work
MDT Total downtime ÷ Number of downtime events The whole stop, including waits
MTBM Operating time ÷ Number of maintenance events (PM and corrective) All maintenance
Inherent availability A_i = MTBF ÷ (MTBF + MTTR) Corrective repair time only
Operational availability A_o = MTBM ÷ (MTBM + MDT) All downtime, including logistics
Failure rate λ = 1 ÷ MTBF (constant failure rate) Random failures
Reliability R(t) = e^(−λt) Constant failure rate only

MTTR is the one to watch. We teach MTTR as hands-on repair time and MDT as the whole stop. SMRP's 7th edition isn't consistent here: its mean-downtime metric and availability guideline treat MTTR as pure repair time, while its MTTR metric starts the clock when the asset stops and includes administrative and logistic delay. Use the definition the question gives. Our MTBF and MTTR pages cover the full availability family.

The reliability formula is standard exponential-model math, not an SMRP metric. One result worth knowing: at t = MTBF, R = e^(−1) ≈ 0.37, so a constant-failure-rate asset has only about a 37% chance of reaching its MTBF without failing.

Where are your gaps? A formula sheet only helps if you know which pillars need the work. Our free 30-question CMRP diagnostic takes about 45 minutes and gives you a score for each of the five pillars.

Pillar 4: Organization and Leadership

Metric Formula
Rework % (Corrective hours identified as rework ÷ Total maintenance labor hours) × 100
Training cost per employee Total formal training cost ÷ Number of internal maintenance employees
Training ROI % [(Business benefits − Training cost) ÷ Training cost] × 100

Pillar 5: Work Management

Metric Formula
Planned work % (Planned hours executed ÷ Total maintenance labor hours) × 100
Reactive work % (Hours of work that broke into the weekly schedule ÷ Total maintenance labor hours) × 100
Schedule compliance (hours) % (Scheduled work performed ÷ Total time available to schedule) × 100
PM & PdM compliance PM/PdM work orders completed by the due date ÷ PM/PdM work orders due
Planned backlog (weeks) (Planned work + Ready work) ÷ Weekly crew capacity
Ready backlog (weeks) Ready work ÷ Weekly crew capacity
Wrench time % (Wrench-time observations ÷ Total observations) × 100
Stores inventory turns Value of stock purchased in the period ÷ Value of stock on hand

Reactive and unplanned aren't the same thing. Reactive work broke into the schedule. Unplanned work never had a job plan.

Worked example: one plant, one month

A packaging line and its maintenance crew over a 30-day month.

Variable key

Variable Meaning Value
T_a Total available time 720 h
T_i Idle time (no orders) 120 h
T_sch Scheduled downtime (PM, changeovers) 24 h
T_un Unscheduled downtime (4 failures) 36 h
T_r Hands-on repair time for the 4 failures 14 h
H_a Crew hours available to schedule per week (8 techs × 40 h) 320 h
H_s Scheduled work performed in a typical week 248 h
W_p Planned work not yet ready (awaiting parts) 900 h
W_r Ready work 700 h
C Weekly crew capacity (no overtime or absences) 320 h
O_w, O_t Work-sampling observations: wrench time, total 128, 400
H_rw, H_m Rework hours, total maintenance labor hours for the month 18 h, 1,280 h

Step 1: Downtime and uptime. T_d = 24 + 36 = 60 h T_o = 720 − (120 + 60) = 540 h

Step 2: Availability vs uptime %. A = 540 ÷ (720 − 120) = 540 ÷ 600 = 90.0% Uptime % = 540 ÷ 720 = 75.0% Same numerator, different denominators, a 15-point gap. This is the most common trap on the sheet.

Step 3: MTBF, MTTR, and MDT. Assume the line ran during all 540 uptime hours. MTBF = 540 ÷ 4 = 135 h MTTR = 14 ÷ 4 = 3.5 h MDT (failure events only) = 36 ÷ 4 = 9.0 h The crew spends 3.5 hours fixing each failure, but the line is down 9. The other 5.5 hours are waiting: a parts and planning problem, not a skills problem.

Step 4: Schedule compliance. SC = 248 ÷ 320 = 77.5%

Step 5: Backlog. Planned backlog = (900 + 700) ÷ 320 = 5.0 weeks Ready backlog = 700 ÷ 320 = 2.2 weeks

Step 6: Wrench time and rework. Wrench time = 128 ÷ 400 = 32.0% Rework = 18 ÷ 1,280 = 1.4%

Dividing rework by corrective hours only would give a bigger, wrong number.

Five denominator traps

  1. Availability vs uptime %: T_a − T_i vs T_a.
  2. Calendar hours in MTBF: use operating time.
  3. Planned vs ready backlog: only planned backlog includes work waiting on parts.
  4. Rework: divide by total maintenance labor hours, not corrective hours.
  5. Cycle time: if performance tops 100%, you inverted it.

A few formulas outside SMRP's metric set

CMRP prep also uses general formulas that aren't numbered SMRP metrics:

Formula Expression
Risk priority number (FMEA) RPN = Severity × Occurrence × Detection, each usually rated 1–10
Simple payback Initial investment ÷ Annual net savings
Net present value NPV = Σ [CF_t ÷ (1 + r)^t] − Initial investment

Variable key: CF_t = net cash flow in year t; r = discount rate; t = year number.

Some sources score RPN from criticality and probability only, so read the question for which factors it uses. Automotive FMEA practice has also moved away from RPN: the 2019 AIAG & VDA FMEA Handbook replaced it with an Action Priority table.

Turning the sheet into recall

Reading a formula sheet won't get you through a closed-book exam. Write one pillar's formulas from memory each day and check them against the sheet. Practice on questions with distractors, not just flashcards. And keep an error log: most misses will be denominators.

The CMRP exam guide covers format and how to prepare, and How hard is the CMRP exam? walks through pacing 110 questions in 2.5 hours.

Our CMRP Practice Exams pack has 600 original practice questions: 3 full-length 110-question practice exams, a diagnostic, and a practice quiz for each of the five pillars, with explanations for every answer.

Find out where you stand

Before you memorize all of this, see which pillars need the most work. The free 30-question CMRP diagnostic is split 6/5/6/5/8 across the five pillars and takes about 45 minutes. You'll get a score for each pillar, so you know where to start.


Sources: SMRPCO Candidate Handbook (posted Oct 2026: https://smrp.org/wp-content/uploads/2026/10/SMRP-Candidate-Handbook.pdf), for the closed-book format, on-screen calculator, and 110 questions in 2.5 hours. SMRP Best Practices, 7th edition, for the metric formulas (1.3–1.5, 2.1.1–2.5, 3.5.1–3.5.5, 4.1, 4.2.1, 4.2.3, 5.3.1, 5.4.1, 5.4.3, 5.4.8, 5.4.9, 5.4.14, 5.5.31, 5.6.1), Guideline 2.0 (OEE), and Guideline 6.0 (availability). The book is inconsistent on MTTR and on the cycle-time performance ratio, as noted above. NIST/SEMATECH e-Handbook of Statistical Methods, section 8.1.6.1, for the exponential reliability formulas. AIAG & VDA FMEA Handbook (2019) for the change from RPN to Action Priority. Payback and NPV are standard engineering-economics definitions. Worked-example figures are invented for illustration.

Independent CMRP exam prep. Not affiliated with SMRP.